Europe knows exactly how much solar it wants. Under the EU’s solar strategy, the bloc is aiming for 700 GW of installed photovoltaic capacity by 2030, and a revised Energy Performance of Buildings Directive now carries a mandate to put panels on new and renovated buildings across the continent. The ambition isn’t the problem. The problem is that nobody, until very recently, could say with any precision where all that solar should actually go.
That sounds like a strange gap for a mature industry to have. Solar in Europe is not short of money, and it is certainly not short of policy. Yet the sector has spent years running on a surprisingly fragmented process at the ground level: developers hunting one rooftop at a time, guessing at which buildings are suitable, and then trying to work out who owns them. The result is a market that knows its destination in aggregate but stumbles building by building. And it is exactly this gap, between macro ambition and micro knowledge, that one geospatial AI startup is now trying to close. One they are accomplishing with a recent investment from a Dubai-based sustainability firm.
Planno, the Geospatial AI (GeoAI) platform for solar developers, today announced a strategic investment from Incubayt Investments, founded by Sami Khoreibi, who built the Arab world’s first utility-scale solar developer, scaled it to $750M+ in revenue across nine countries, and exited it to a UK pension fund. The investment is Incubayt’s first in geospatial AI for solar, and it carried Planno from Dubai into 16 markets.
“We invest in founders building globally ambitious projects from the UAE,” said Sami Khoreibi.
“The solar industry has never been short of capital or sunlight. What it lacked was intelligence. Where to build, and who owns the roof. In a data-driven economy, that insight is as valuable as the projects themselves. Planno is exactly that.”
The scale of what’s sitting unused
In January 2026, the European Commission’s Joint Research Centre, its in-house science service, published something that had never existed before: a per-building estimate of rooftop solar potential across the entire EU. Drawing on a high-resolution model of all 271 million buildings, the researchers concluded that Europe’s rooftops could hold roughly 2.3 terawatts of solar capacity and generate about 2,750 terawatt-hours a year. That’s close to 40% of the electricity a fully renewable EU would need by 2050.
However, only around 10% of European building rooftops currently have any solar on them at all. Rooftop systems already punch well above their weight, making up about 61% of the EU’s installed PV capacity, but the vast majority of viable roofs remain bare. The JRC was blunt about the commercial opportunity in particular: large non-residential buildings over 2,000 m² alone could host around 355 GW, and non-residential rooftops together could cover more than half of the entire 2030 target.
Those are enormous numbers hiding in plain sight, on warehouses, factories, logistics parks and retail sheds that already exist. The barrier was never really the physical roof. It was the intelligence needed to identify them until now.
Why mapping intelligence became the constraint
For most of solar’s growth, the binding constraints were the obvious ones, including panel costs, financing, grid connections, and permits. Those haven’t disappeared, but they’ve eased enough that a subtler bottleneck has surfaced. A developer trying to build a pipeline of commercial rooftop projects has to answer a deceptively hard sequence of questions for every candidate building: Is the roof structurally and geometrically suitable? How much capacity could it hold? Is it already fitted? And crucially, who owns it, and how do I reach them?
Historically, this was manual work, done more or less blind, and it didn’t scale. A firm might identify a few good sites a week by squinting at satellite imagery and cross-referencing property records by hand. Alternatively, like many, they could purchase costly leads that are also shared amongst their competitors.
Multiply that friction across a continent of 271 million buildings, and you begin to see why adoption has been so lopsided. Including why the roofs that do get built tend to be the biggest, most obvious ones, while everything mid-sized gets left behind.
The interesting shift is that this is a problem new development in software is unusually well suited to solve. Satellite and aerial imagery are now detailed enough to assess a roof remotely. Machine-learning models can classify surfaces, estimate usable area and flag which buildings already have panels. And that imagery can be fused with property and market data to attach an owner and a rough economic case to each roof.
What used to be weeks of legwork becomes a filterable list. This is the premise behind the JRC’s own open dataset, which it explicitly pitched as a tool for “policymakers, planners, utilities, and investors” to plan deployment more precisely.
The startup chasing the gap
One of the more visible names here is Planno, a geospatial-AI platform that maps commercial and industrial rooftops, estimates their solar potential, and identifies their owners so developers can prospect at scale rather than one building at a time.
Founded by Portuguese energy engineer Daniel Domingues and headquartered in Dubai, it recently raised backing from Incubayt Investments and has since expanded into markets across Europe, the Middle East, Asia, Africa and North America.
Planno’s own market analyses read like a street-level version of the JRC’s continental one. In Switzerland’s Lake Geneva region, the company found that while more than half of the very largest rooftops already carried solar, adoption across all commercial and industrial buildings sat at just 24%, the familiar pattern of the biggest owners moving first and the long tail lagging.
Domingues frames the value of this kind of intelligence in economic terms, with electricity demand climbing and prices increasingly volatile; he argues rooftop solar is “one of the few cost-control levers a commercial operator actually owns.” That’s a notably different approach from the subsidy-driven rhetoric of solar’s first decade, and it may prove more durable as we step into a future in which energy looks different.
The broader field of solar and rooftop intelligence includes players building similar layers from remote roof-assessment tools used by installers to large-scale potential-mapping projects run by researchers and municipalities. What unites them is a recognition that the next phase of solar growth is less about generating power than about generating information about where power should be generated.
Why this matters beyond solar
There’s a wider lesson here about how energy transitions actually happen. We tend to imagine them as a story of hardware—cheaper panels, better batteries, bigger turbines. But hardware alone doesn’t deploy itself. Somewhere between a national target and a physical installation sits an enormous amount of unglamorous coordination: knowing what exists, what’s viable, and who to call. When that coordinating knowledge is missing, ambition stalls in exactly the way Europe’s rooftop numbers reveal: a 90% gap between potential and reality, despite every incentive pointing the same direction.
This is why the quiet arrival of building-level data, whether from a public institution like the JRC or a commercial platform, may end up mattering as much as the mandates themselves. A solar mandate tells the market it must build. A map tells it where, and turns a policy obligation into an actual, prioritised list of projects.
